Forum
PLDX
Created 24th January 2011 @ 11:10
Add A Reply Pages: 1 2 Next »
When i try using PLDX it laggs out my tf2, well it doesnt lag it out it…goes really slowly, i mean like host_timescale 0.25 slow or slower and i cant get it to speed up or anything. any advice? or is pldx never gonna work for me?
Last edited by Bonkers,
Quoted from octochris
that’s normal when recording.
really? when i play it in Windows media player it’s the same speed though =S
Quoted from Bonkers
[…]
really? when i play it in Windows media player it’s the same speed though =S
it’s obviously going to lag if you’re playing uncompressed footage straight from your hard drive…
Quoted from octochris
[…]
it’s obviously going to lag if you’re playing uncompressed footage straight from your hard drive…
meh idk shit bout this crap anyway so, sorry for the trouble =P
edit: Bloody typos
Last edited by Bonkers,
The higher the fps you record at the slower it will run. Download Virtual Dub from here: http://virtualdub.sourceforge.net/
After you record your clip, open VDub and open file, go to the a_000001.tga file and click it. You should see the start of your clip.
Now go to audio, add audio from another file and select the a1_wav file. That will sync the audio and video together.
Choose video, frame rate and select the same fps as you want to render at, 30 probably.
Now the slightly complicated bit, decimate by the recording fps/render fps. So if you record at 120 and render at 30 you decimate by 4.
Next choose compression and use the lagarith lossless codec available here: http://lags.leetcode.net/codec.html
Then go to save as avi., give it a name and wait while it sorts it out.
Upload clip to Vegas/WMM/AE, edit and play with it, render it from there and away you go.
Any questions add me on steam and I can help you out..if my home web ever gets fixed…… :(
Ermmmm, is that for real monkeh or are you trolling?
I very rarely troll intentionally, that’s how I do it. Maybe it’s long winded and stoopid, but it works for me. Granted, maybe I should just record at 30 fps and be done with it, in fact I probably will from now on, but no, I wasn’t trolling, failing maybe, but not trolling.
What monkey said isn’t far off, only thing i would change is when you set the framerate in Virtual Dub set it to the same framerate you recorded at (eg 120)
If I do that BoneS then I get a horrible blurry effect.
Here’s a vid I made using 120 FPS and no decimation: http://www.youtube.com/watch?v=kPVzJk2JUVM see how it’s all fuzzy around the edges with any kind of movement? I decided that was because VDub was cramming 4 capture frames into one frame of dub.
Using the decimation and having all other settings the same I get this: http://www.youtube.com/watch?v=ONZNaeRPiIk which is a lot smoother.
However it does show that 120 isn’t enough for decent slo mo :/ But then saying that if I record at 480 the player models still jump around and skip, even with cl_smooth and cl_smoothtime.
Quoted from BoneS
What monkey said isn’t far off, only thing i would change is when you set the framerate in Virtual Dub set it to the same framerate you recorded at (eg 120)
Nop. Leave it to Automatically find the FPS. I explained why in the recording tutorial I made: http://www.youtube.com/watch?v=CeU0M4XaL4k
Bassically, TF2 never records at bang on the FPS you tell it to, so if you manually give it the FPS then you will end up with the audio being de-synced from the video.
I’ve never noticed my audio de-syncing tbh, but I’ve never recorded anything over 30-40 seconds or so! Hmm, brb, watching tutorial…
Last edited by Monkeh,
Sorry Monkeh, I should have been more specific. I was referring to the decimation part of your post. As you later say, it would be just as easy (and faster) to record at 30fps.
The ‘fuzzy edge’ you refer to later is more commonly called frame-blending, whereby, (I’m sure you already know all this) several instances of the ‘same’ frame are composited in such a way to create the illusion of motion. The only reason yours is so ‘fuzzy’ is because you didn’t over-sample enough times.
The reverse is true for your Jamie T-(irrel) intro. Because you decimated frames there is no oversampling, thus making it appear sharper. The downside in this instance is you do notice the lack of motion-blur when the demoman shoots offscreen.
Please take into account when replying to me that I am incredibly idiotic, hugely inexperienced and often referred to as ‘a retard’! Suggesting I already know anything is never a safe bet, please treat me like a 4 year old and explain everything nice……and…..slowly :)
The way I record is a spliced together version of how to do things that I worked out myself. I read posts and tutorials and came up with that method…after quite a lot of experimenting.
I had no previous experience in movie making, (I know, hard to believe isn’t it?!….Ahem), or any kind of editing, image capture, demo capture or anything that may have been useful, so I started from scratch, made a shitty video using the free version of fraps: http://www.youtube.com/watch?v=E8zHdlbMaSg (so so so so bad), and then decided it was fun and I wanted to make better looking material, which I think I’ve managed to some degree.
So, sampling and frame blending? What is all this please and how does one get super shiny videos the like of which will make you weep at the sheer beauty?
Well, in this case, over-sampling is recording at a higher fps than your final movie fps (ie, recording at 120fps and rendering your final movie at 30fps means 120/30=4 samples per frame).
Frameblending four samples per frame is not really high enough for high-motion video (such as fragmovies) and is the reason your Badlands movie looks ‘fuzzy’ as you can clearly see the four samples when they are frameblending in high-motion scenes (ie when he looks around quickly).
Frameblending is an attempt to create the illusion of motion-blur by overlaying those extra frames you have recorded and changing their opacity. AE uses a pretty crude method of incrementally reducing the opacity of the trailing edge frames (I don’t know anything about Vegas, but I’m guessing it does something pretty similar). VirtualDub does not do any frameblending. There are other apps that do a much much better job of creating motion-blur, and there are also plugins (for AE at least) that will do a very good job too.
Eight frameblended samples per frame is the bare-minimum for low-motion video, so I would say 16 samples would be the bare-minimum for fragmovies, requiring you to record at 16*30=480. As a sidenote, that ValveFilmMaker(?) thing uses 64 samples to create its motion-blur (64*30=1920!!!)
As for your last question, the answer is easy. You hit the MAKE PRETTY button, and let the ‘computers do all the work’ of course.
Add A Reply Pages: 1 2 Next »
content rss